Paper

When to Trust Adaptation: Contextual Fusion for Transferable Vision-Language Prompting · ByteTrail